One of the hallmarks of tumor cells is genetic rearrangements, including deletions, translocations, inversions, and mitotic recombination. In many cases these lesions result in the activation of oncogenic mutations or loss of heterozygosity (LOH) of tumor suppressor genes. The initiating lesion for these types of recombination is DNA strand breaks (590). Thus the generation of strand breaks could well be a necessary step for the evolution of some tumor types, although it is not clear to what extent this might be rate limiting for tumorigenesis. Loss of p53 function is another very common genetic alteration in tumors. In normal cells p53 is induced in response to DNA damage, specifically dsbs, resulting in cell cycle arrest and/or apoptosis of the damage cell. Loss of p53 would render the cells permissive to accumulate additional genetic lesions, for example dsb-induced rearrangements. Severe combined immunodeficient (SCID) mice are severely defective in repair of dsbs and are sensitive to spontaneous and radiation-induced lymphomagenesis. We suggest that these mice represent a very useful model to examine the role of dsbs in tumorigenesis of several different tissues. They will also be useful to dissect the molecular pathway from dsbs to p53 induction to apoptosis in a tissue specific context. Aim one will determine if scid/scid and scid/+ mice are susceptible to several tissue specific carcinogenesis protocols including skin, liver, and lung. This will address the question of the importance of dsbs in the evolution of different tumor types. It will also reveal if heterozygous carriers of this defect are at increased risk. If so this could have implications for human heterozygous carriers of similar defects. Aim two will examine the susceptibility of scid/scid mice to specific dsb-inducing chemotherapeutic agents including etoposide, campothecin and bleomycin. This will determine if dsb induction can be a rate limiting step for carcinogenesis. Aim three will determine if the p53 induction, apoptosis pathway is abnormal in SCID mice. Aim three will also determine if radiation-induced p53 expression and apoptosis differ substantially between tumor types (skin, liver and lung) and if p53 genotype predicts the apoptotic response of tumor cells in vivo. Aim four will examine mutational inactivation of p53 in tumors induced in SCID mice to determine if there is unusually strong selection against p53 function. Allelotype analysis in aim five will be used to determine if the frequency of large scale genetic rearrangements, such as deletions, is increased in tumors from SCID mice, as well as to identify alternative genetic pathways which might complement p53 inactivation.